JACC: CARDIOVASCULAR IMAGING VOL. 6, NO.

10, 2013

ª 2013 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION ISSN 1936-878X

PUBLISHED BY ELSEVIER INC. THIS IS AN OPEN ACCESS ARTICLE UNDER THE http://dx.doi.org/10.1016/j.jcmg.2013.03.008

CC BY-NC-ND LICENSE (http://creativecommons.org/licenses/by-nc-nd/4.0/).

T1 Mapping for the Diagnosis of

Acute Myocarditis Using CMR
Comparison to T2-Weighted and Late Gadolinium Enhanced Imaging

Vanessa M. Ferreira, MD, DPHIL,* Stefan K. Piechnik, PHD, MSCEE,*

Erica Dall’Armellina, MD, DPHIL,* Theodoros D. Karamitsos, MD, PHD,*
Jane M. Francis, DCR(R), DNM,* Ntobeko Ntusi, MB, CHB,* Cameron Holloway, DPHIL,*
Robin P. Choudhury, DM,* Attila Kardos, MD, PHD,y Matthew D. Robson, PHD,*
Matthias G. Friedrich, MD,zx Stefan Neubauer, MD*
Oxford and Milton Keynes, United Kingdom; and Calgary, Alberta, and Montréal, Quebec, Canada

OBJECTIVES This study sought to test the diagnostic performance of native T1 mapping in acute
myocarditis compared with cardiac magnetic resonance (CMR) techniques such as dark-blood T2-
weighted (T2W)-CMR, bright-blood T2W-CMR, and late gadolinium enhancement (LGE) imaging.

BAC KG R O U N D The diagnosis of acute myocarditis on CMR often requires multiple techniques,
including T2W, early gadolinium enhancement, and LGE imaging. Novel techniques such as T1 mapping
and bright-blood T2W-CMR are also sensitive to changes in free water content. We hypothesized that
these techniques can serve as new and potentially superior diagnostic criteria for myocarditis.

M E T H O D S We investigated 50 patients with suspected acute myocarditis (age 42  16 years; 22%

women) and 45 controls (age 42  14 years; 22% women). CMR at 1.5-T (median 3 days from presenta-
tion) included: 1) dark-blood T2W-CMR (short-tau inversion recovery); 2) bright-blood T2W-CMR (acqui-
sition for cardiac unified T2 edema); 3) native T1 mapping (shortened modified look-locker inversion
recovery); and 4) LGE. Image analysis included: 1) global T2 signal intensity ratio of myocardium
compared with skeletal muscle; 2) myocardial T1 relaxation times; and 3) areas of LGE.

R E S U LT S Compared with controls, patients had significantly higher global T2 signal intensity ratios by
dark-blood T2W-CMR (1.73  0.27 vs. 1.56  0.15, p < 0.01), bright-blood T2W-CMR (2.02  0.33 vs. 1.84
 0.17, p < 0.01), and mean myocardial T1 (1,010  65 ms vs. 941  18 ms, p < 0.01). Receiver-operating
characteristic analysis showed clear differences in diagnostic performance. The areas under the curve for
each method were: T1 mapping (0.95), LGE (0.96), dark-blood T2 (0.78), and bright-blood T2 (0.76). A T1
cutoff of 990 ms had a sensitivity, specificity, and diagnostic accuracy of 90%, 91%, and 91%, respectively.

CONCLUSIONS Native T1 mapping as a novel criterion for the detection of acute myocarditis
showed excellent and superior diagnostic performance compared with T2W-CMR. It also has a higher
sensitivity compared with T2W and LGE techniques, which may be especially useful in detecting
subtle focal disease and when gadolinium contrast imaging is not feasible. (J Am Coll Cardiol Img
2013;6:1048–58) ª 2013 by the American College of Cardiology Foundation. Published by Elsevier Inc.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).

From the *Department of Cardiovascular Medicine, University of Oxford, John Radcliffe Hospital, Oxford, United Kingdom;
yDepartment of Cardiology, Milton Keynes NHS Hospital Foundation Trust, Milton Keynes, United Kingdom; zStephenson
Cardiovascular MR Centre, Libin Cardiovascular Institute of Alberta, Calgary, Alberta, Canada; and the xDepartment of
Cardiology, Université de Montréal, Montréal, Quebec, Canada. This study is funded by the Oxford National Institute for
Health Research Biomedical Research Centre Programme. Dr. Ferreira is funded by the Alberta Innovates Health Solutions
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013 Ferreira et al. 1049
OCTOBER 2013:1048–58 ShMOLLI T1 Mapping Detects Myocarditis

A
cute myocarditis is common, accounting for
up to 75% of patients presenting with chest
pain, raised troponin, but nonobstructive
coronary arteries (1–4); 12% of young adults
presenting with sudden death (5); and 9% of patients
who develop dilated cardiomyopathy (6). It is chal-
lenging to diagnose, often requiring a combination of
clinical assessment and diagnostic tests. Cardiac
magnetic resonance (CMR) is the imaging tool of
choice because of its ability for multiparametric tissue
characterization, as established in recent consensus
guidelines (7).
See page 1059
Currently, 3 CMR methods are available to assess
different pathophysiological processes in acute
myocarditis: dark-blood T2-weighted (T2W) imag-
ing for edema, T1-weighted imaging pre- and post-
contrast for hyperemia, and late gadolinium
blood T2W-CMR in acute myocarditis compared
with dark-blood T2W-CMR and LGE imaging.
METHODS
Study population. This was a prospective study
enrolling consecutive patients (n ¼ 50) with sus-
pected myocarditis (7,13) from 2 hospitals (1 tertiary
care centerdThe John Radcliffe Hospital, Oxford,
United Kingdomdand 1 medium-sized district
general hospitaldMilton Keynes Hospital, Milton
Keynes, United Kingdom). Patients underwent
CMR scanning at the John Radcliffe Hospital be-
tween January 2010 and March 2012. All patients
had: 1) acute chest pain; 2) elevation in cardiac
troponin I level (>0.04 mg/l); and 3) history of recent
systemic viral disease or absence of significant
(>50%) obstructive coronary artery disease
ABBREVIATIONS
on coronary angiography or absence of risk
AND ACRONYMS
factors for coronary artery disease or age
<35 years. Exclusion criteria included AUC = area under the curve

enhancement (LGE) to detect patterns of myocyte contraindications to CMR, previous CMR = cardiac magnetic
resonance
necrosis (7). Although CMR is a powerful tool, there myocardial infarction, previous myocarditis,
are some recognized limitations related to technical or any chronic cardiac conditions. Patients EGE = early gadolinium
enhancement
factors, image analysis, and the mechanism of each who demonstrated myocardial infarction as
EMB = endomyocardial biopsy
method to detect myocarditis (7,8). Using a com- evidenced by an ischemic pattern of LGE
LGE = late gadolinium
bined CMR approach with at least 2 of these 3 criteria (i.e., an isolated area involving the sub-
enhancement
can improve diagnostic accuracy (7). endocardium) or an obvious alternative
PPV = positive predictive value
Novel CMR techniques are constantly being diagnosis on CMR (such as Takotsubo or
ROC = receiver-operating
developed, some of which may be suitable for the hypertrophic cardiomyopathy) were also characteristic
evaluation of myocarditis. For instance, bright- excluded. Healthy volunteers (n ¼ 45) with
ShMOLLI = shortened modified
blood T2W has been shown to be superior to no cardiac history or known cardiac risk look-locker inversion recovery
dark-blood T2W imaging for detecting the area at factors, not on cardiovascular medications, SI = signal intensity
risk in acute myocardial infarction (8–10). Quanti- and with a normal electrocardiogram un- STIR = short-tau inversion
tative techniques such as T1 mapping allow direct derwent CMR as controls. All subjects gave recovery
tissue characterization and have been shown to be written informed consent, and ethical T2W = T -weighted 2

superior to T2W techniques in detecting acute
edema (11) and area at risk in acute myocardial
infarction (12). These methods are promising but
have not been systematically assessed in acute
myocarditis.
In this prospective study, we performed multi-
parametric tissue characterization in patients with
suspected acute myocarditis (7) using novel and
conventional CMR techniques. We aimed to test the
diagnostic performance of T1 mapping and bright-
approval was granted for all study
procedures.
Cardiac magnetic resonance. CMR studies were
performed within 14 days of symptoms using a single
1.5-T magnetic resonance system (Avanto, Siemens
Healthcare, Erlangen, Germany). CMR included
cine, T2W, T1 mapping, and LGE imaging, with
matching short-axis images. T1 mapping was per-
formed using the shortened modified look-locker
inversion recovery (ShMOLLI) sequence (14);

Clinical Fellowship and the University of Oxford Clarendon Fund Scholarship. Dr. Choudhury is a Wellcome Trust Senior Research Fellow
in Clinical Science. Drs. Choudhury and Neubauer acknowledge support from the British Heart Foundation Centre of Research Excellence,
Oxford. Drs. Piechnik and Robson have patent authorship rights for U.S. patent pending 61/387,591. SYSTEMS AND METHODS FOR
SHORTENED LOOK LOCKER INVERSION RECOVERY (Sh-MOLLI) CARDIAC GATED MAPPING OF T1. September 29, 2010.
Dr. Friedrich is a board member, advisor, and shareholder of Circle Cardiovascular Imaging Inc. All other authors have reported that they
have no relationships relevant to the contents of this paper to disclose. Drs. Ferreira and Piechnik are joint first authors of this work. Drs.
Neubauer and Friedrich are joint senior authors of this work.
Manuscript received March 6, 2013; accepted March 29, 2013.
1050 Ferreira et al.
ShMOLLI T1 Mapping Detects Myocarditis
dark-blood and bright-blood T2W-CMR were
performed with the short-tau inversion recovery
(STIR) (7,15) and the acquisition for cardiac unified
T2 edema (9) sequences, respectively. All were ac-
quired before administration of contrast agents. LGE
imaging was acquired using a T1-weighted phase-
sensitive inversion recovery sequence (16) 8 to 10
min after intravenous administration of contrast
agent (gadodiamide [Omniscan], GE Healthcare,
Chalfont St. Giles, United Kingdom) (total 0.13
mmol/kg body weight at 6 ml/s). A 32-channel
phased-array chest coil was used for all data acquisi-
tion, except for STIR imaging for which the body
coil was used. Acquisition parameters are provided
in the Online Appendix.
Image analysis. Matching short-axis slices were
compared across cine, dark-blood T2W-CMR,
bright-blood T2W-CMR, T1 mapping, and LGE
imaging. Analysis of left ventricular ejection fraction
was performed using Argus software (Siemens
Healthcare). Short-axis images from all methods
were manually contoured using in-house software
MC-ROI (programmed by S.K.P. in Interactive
Data Language, version 6.1, Exelis Visual Informa-
tion Solutions, Boulder, Colorado) to outline the
endocardium and epicardium, and then divided into
6 segments per slice using the anterior right ventric-
ular–left ventricular insertion point as reference and
for comparing segments among sequences. Analysis
of wall motion, T2W images, LGE images, and T1
mapping was performed as previously published
(7,11), with additional details provided in the Online
Appendix. For all quantitative analysis of T2W,
LGE, and T1 map images, only regions of myocar-
dium with a contiguous area of $40 mm2 above the
specified thresholds were considered relevant. This
corresponds to 10 adjacent pixels for the STIR
method, in accordance with currently proposed rec-
ommendations (7), to reduce the detection of noise as
positive findings. To determine the sensitivity and
specificity for each threshold, average involvement of
at least 5% of at least 1 myocardial segment per subject
above the specified threshold was considered a posi-
tive finding by each method.
Statistical analysis. Normality of data was tested us-
ing the Kolmogorov-Smirnov test. Normally
distributed data are presented as mean  SD;
nonparametric data are presented as medians with
interquartile ranges. Unpaired samples between
groups were assessed by the unpaired 2-tailed Student
t test, or the Mann-Whitney U test for nonparametric
data. Correlation was assessed using the Pearson r and
Spearman rho coefficient as appropriate. Segmental
data were averaged on a per-subject basis before any
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013
OCTOBER 2013:1048–58
interindividual and group comparisons to control for
clustering of segments within each subject. Receiver-
operating characteristic (ROC) analysis was per-
formed to compare the diagnostic performance of the
CMR methods in detecting myocardial changes in
patients compared with controls. Significance of the
ROC analyses was assessed using the method of
Delong et al. (17) (MedCalc, version 11.5.1.0,
MedCalc Software, Mariakerke, Belgium). The
McNemar test and Cochran’s Q test were used for
comparing combinations of CMR tissue character-
ization methods in the diagnosis of acute myocarditis.
All statistical tests were 2-tailed, with p values <0.05
considered statistically significant. To determine the
presence of significant differences in subject groups
when using multiple CMR methodologies, analysis
of variance was performed with Bonferroni-corrected
post hoc comparisons for parametric data; for
nonparametric data, the Kruskal-Wallis 1-way anal-
ysis of variance was performed with post hoc pairwise
comparisons using the Mann-Whitney U test. The
p value for significance was adjusted from p < 0.05 to
p < 0.01 for multiple comparisons within groups
where appropriate.
RESULTS
Detection of myocardial changes by CMR in acute
myocarditis. Table 1 provides the study patient
characteristics. Controls were sex-matched and of
similar age distribution (22% women; age 42  14
years). CMR findings are shown in Figure 1 and
summarized in Table 2. Patients had significantly
lower mean left ventricular ejection fractions
compared with controls (ejection fraction 63  12%
vs. 72  6%; p < 0.01). Patients had significantly
higher T2 signal intensity (SI) ratios, as measured by
both dark-blood T2W and bright-blood T2W
imaging, as well as significantly higher mean
myocardial T1 values and LGE SI ratios compared
with controls. The pattern of LGE was predomi-
nantly subepicardial (95.6%) and midwall (84.4%),
localized most frequently to the lateral wall (97.9%)
and inferior wall (95.6%), as compared with
the septum (60.0%) and anterior wall (35.6%). None
of the 50 patients demonstrated an isolated
ischemic (subendocardial) pattern of LGE to suggest
myocardial infarction as the etiology of the presen-
tation. Five of 50 patients (10%) had a subend-
ocardial LGE pattern in conjunction with other
patterns typical for myocarditis.
Relationship between myocardial T1 and other measures
of acute myocardial injury. Within the patient group,
average myocardial T1 values correlated well with
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013
OCTOBER 2013:1048–58
Table 1. Baseline Characteristics of the Patient Cohort
(n [ 50)
Age, yrs
Female
Risk factors
Hypertension
Diabetes
Hyperlipidemia
Smoking
Family history of premature CAD
Presenting characteristics
Recent viral illness
Recent inflammatory illness
Troponin I, mg/l 5.15 [1.22 to 14.20]
C-reactive protein, mg/l
Coronary angiography
Nonobstructive coronaries on angiography
Coronary angiogram not performed
Young age, <35 yrs
No risk factors for CAD, age >35 yrs
Recent viral/inflammatory illness,
age >35 yrs
Time from symptoms to CMR, days
Values are n, n (%), or median [interquartile range].
CAD ¼ coronary artery disease; CMR ¼ cardiac magnetic resonance.
left ventricular ejection fraction (r ¼ 0.57;
p < 0.0001) and better than measures of global
myocardial edema by dark-blood T2W (r ¼ 0.34;
p < 0.02) or bright-blood T2W imaging (r ¼ 0.44;
p < 0.003). Mean myocardial T1 showed moderate
correlations with relative T2 SI changes by dark-
blood T2W, bright-blood T2W, and LGE imag-
ing (r ¼ 0.51, 0.44, and 0.49, respectively; all
p < 0.003). There were significant correlations be-
tween troponin I levels with focal areas of myocardial
injury as measured by T1 >990 ms (r ¼ 0.54;
p < 0.0001), LGE (r ¼ 0.43; p < 0.002), dark-blood
T2W (r ¼ 0.62; p < 0.0001), and bright-blood
T2W imaging (r ¼ 0.37; p ¼ 0.01). T1 values were
significantly higher in patient myocardial tissue
characterized by an increased dark-blood T2 SI ratio
(1,055  63 ms), increased bright-blood T2 SI ratio
(1,032  70 ms), and LGE (1,045  66 ms),
compared with myocardial tissue of controls (941 
18 ms; all p < 0.01). For “CMR-negative” myocar-
dial tissue in patients, that is, those without LGE
and no significantly increased T2 SI ratios on both
dark-blood and bright-blood T2W imaging, T1
values were also significantly higher compared with
controls (978  45 ms vs. 940  28 ms; p < 0.001).
Ferreira et al. 1051
ShMOLLI T1 Mapping Detects Myocarditis
Diagnostic performance of CMR methods in the
detection of acute myocarditis. To maximize the
42  16
accuracy of the threshold values to detect acute
myocarditis using these methods, each segment was
11 (22)
strictly assessed for image quality before inclusion
into the final ROC analysis, and only segments with
11 (22)
no or minimal artifacts were included, as previously
3 (6) published (11). On cine imaging, 3.5% of segments
7 (14) corresponding to the left ventricular outflow tract
16 (32) were rejected; on dark-blood STIR imaging, 8.6%
5 (10) were rejected because of artifacts, signal dropout, or
breathing motion; and on bright-blood acquisition
41 (82)
for cardiac unified T2 edema imaging, 3.7% of seg-
ments were rejected because of off-resonance or pa-
5 (10)
tient movement artifacts. T1 maps were assessed for
quality in 3 ways: examination of the T1 maps, the
21 [4 to 89]
raw T1 images, and R2 maps; 11.2% of the segments
were excluded because of off-resonance artifacts, poor
36 (72) T1 fit on the R2 maps, patient movement, or a low
14 (28) signal-to-noise ratio (11). On LGE imaging, 2%
10 were rejected because of artifacts caused by patient
2 movement or poor image quality. No subject was
2
excluded from the final analysis owing to image
quality.
3 [1 to 6] Both T1 mapping and LGE imaging demon-
strated excellent and equivalent diagnostic perfor-
mance, with areas under the curve (AUC) of 0.96
and 0.95, respectively (p ¼ ns). T1 mapping
significantly outperformed dark-blood T2W (AUC:
0.78) and bright-blood T2W imaging (AUC: 0.76;
both p < 0.001). To check for bias against any of
the methods, ROC analysis was repeated using all
data including segments affected by artifacts. All of
the aforementioned relative relationships were pre-
served, with no significant change in the AUC for
each method (all AUC: #0.02).
The diagnostic performance of each CMR tissue
characterization method and their combinations are
summarized in Table 3. T1 mapping, using a cutoff
of T1 $990 ms as established in our previous study
for detecting acute edema (11), demonstrated very
good performance, with a sensitivity, specificity,
diagnostic accuracy, positive predictive value (PPV),
and negative predictive value of w90% across the
board. The T2W methods showed good diagnostic
performance within ranges published in the litera-
ture (7), whereas LGE had a high specificity (97%)
and PPV (97%) with a lower sensitivity (74%)
compared with T1 mapping.
The combination approach represents a tradeoff
between sensitivity and specificity. Combining the
T2W methods with LGE significantly improved
specificity and PPV, with some loss in sensitivity,
thus LGE aided T2W imaging in arriving at a
1052 Ferreira et al. JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013

ShMOLLI T1 Mapping Detects Myocarditis OCTOBER 2013:1048–58

Figure 1. Acute Myocarditis

(A) Dark-blood T2-weighted (T2W) imaging demonstrating increased signal intensity in the mid lateral wall (arrows). (B) Bright-blood T2W
imaging demonstrating increased signal intensity in the mid lateral wall (arrows). (C) Shortened modified look-locker inversion recovery
(ShMOLLI) T1 map demonstrating increased T1 values (1,100 to 1,200 ms) in the lateral wall (arrows). (D) Late gadolinium enhancement (LGE)
imaging demonstrating mid-wall enhancement in the lateral wall (arrows).

positive diagnosis of myocarditis. On the other
hand, compared with T1 mapping alone, combining
T1 mapping with LGE only improved specificity
slightly (from 91% to 97%), but at the expense of
lower sensitivity (dropping from 90% to 70%), and
this combination was not better than LGE alone
(not surprising, given the 2 methods had a very
similar ROC curve, as shown in Fig. 2). However,
the overall diagnostic performance of the T1 þ
LGE combination fared better than any of the
T2W þ LGE combinations. In fact, T1 mapping
Table 2. CMR Findings in Patients With Acute Myocarditis Compared With Controls
CMR Measures Controls
Ejection fraction, % 72  6
Dark-blood T2 SI ratio 1.56  0.15
Bright-blood T2 SI ratio 1.84  0.17
Myocardial T1, ms 941  18
LGE, % myocardium 0% [0% to 2%]
Values are mean  SD or median [interquartile range]. Dark-blood T2 SI ratio and bright-blood T2 SI ratio ¼
SImyocardium/SIskeletal muscle; LGE (%) ¼ median volume fraction of late gadolinium enhancement (LGE) per
subject. *p < 0.01.
CMR ¼ cardiac magnetic resonance; SI ¼ signal intensity.
showed a better diagnostic performance than the
T2W methods, either alone or in combination with
LGE.
DISCUSSION
In this work, we have demonstrated for the first
time to our knowledge that quantitative CMR by
T1 mapping had an excellent diagnostic perfor-
mance in detecting changes in acute myocarditis.
Further, T1 mapping outperformed both dark-
blood and bright-blood T2W techniques, either
alone or in combination with LGE, in the diagnosis
of acute myocarditis, with superior sensitivity and
Patients excellent specificity and diagnostic accuracy.
63  12* The CMR methods tested in this study all have
1.73  0.27*
advantages and limitations. From the technical
2.02  0.33*
standpoint, factors that lead to the failure of a
method to diagnose acute myocarditis may be cate-
1,010  65*
gorized into: 1) compromised image quality during
11% [5% to 21%]*
acquisition (patient movement, poor breath-holding,
tachyarrhythmia); 2) suboptimal post-processing
techniques; and 3) noise and the selection of
thresholds.
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013 Ferreira et al. 1053
OCTOBER 2013:1048–58 ShMOLLI T1 Mapping Detects Myocarditis

Table 3. Diagnostic Performance of CMR Tissue Characterization Methods in the Detection of Suspected Acute Myocarditis

Sensitivity Specificity Accuracy PPV NPV

Individual
T1 mapping* 90 91 91 92 89
Dark-blood T2y 52 84 67 79 61
Bright-blood T2z 67 55 64 78 42
LGEx 74 97 83 97 69
Combination
T1 mapping and LGEk 70 97 80 97 66
Dark-blood T2 and LGE 48 97 66 96 53
Bright-blood T2 and LGE 50 100 55 100 18

Values are %. *Myocardial injury is detected when T1 is $990 ms; yedema is diagnosed when the T2 SI ratio (T2 SImyocardium:skeletal muscle) is $2.0; zedema is diagnosed
when the T2 SI ratio (T2 SImyocardium:skeletal muscle) is $2.2 (equivalent to >2 SD of normal; see Table 2); xLGE is detected when myocardial SI is $2 SD above mean SI
of normal myocardium. kThe combination of T1 mapping and LGE was significantly better in performance than dark-blood T2 þ LGE and bright-blood T2 þ LGE (all
p < 0.005). For each technique, only contiguous areas of myocardium $40 mm2 above the stated threshold were considered relevant; involvement of $5% of
any segment on a per-subject basis was the threshold used for comparison of methods.
NPV ¼ negative predictive value; PPV ¼ positive predictive value; T2W ¼ T2-weighted; other abbreviations as in Table 2.

Dark-blood T2W imaging. Dark-blood T2W-CMR T1 mapping is consistent with skeletal muscle
is widely used for clinical edema imaging. In addi- inflammation demonstrated in patients presenting
tion to following recommended parameters (7), with acute myocarditis using pre-and post-contrast
optimal image quality relies on a regular heart rhythm T1-weighted imaging (18). Thus, in some of our
with a ventricular rate within a relatively normal patients, conventional T2W imaging was unable to
range, appropriate selection of the repetition time diagnose edema in myocarditis even when using
tailored to the patient’s heart rate, and the ability of skeletal muscle as a reference, with the remainder of
the patient to breath-hold. Excessive tachycardia and the patients potentially affected by the same mecha-
irregular rhythms, such as atrial fibrillation, and poor nism, only to a lesser degree, as supported by the
breath-holding render dark-blood T2W images
nondiagnostic most of the time. Dark-blood T2W
images are known for signal dropout, especially in the
basal inferolateral wall, where myocarditis typically
manifests. When myocarditis is global, the use of
“normal remote” myocardium as a reference, which
may not be available, leads to significant underesti-
mation of the extent of myocardial injury (Fig. 3).
The use of skeletal muscle circumvents this problem
unless skeletal muscle is also inflamed, which occurs
in some cases of myocarditis (18) and also in this work
(Fig. 4). This is supported by the findings that pa-
tients with myocarditis demonstrated significantly
higher T1 values in skeletal muscle and with greater
variability (822  55 ms) compared with controls
(803  32 ms; p ¼ 0.04). In this cohort of 50 patients,
there were 3 cases in which the dark-blood T2 edema
ratio relative to skeletal muscle was <2.0 (1.78 
0.11) despite reasonable image quality, but the
average myocardial T1 was significantly increased
(1,158  11 ms) (Fig. 4). This can be explained by
skeletal muscle inflammation, supported by signifi-
Figure 2. Diagnostic Performance of T1 Mapping Compared With T2W and
cantly increased T1 values in the skeletal muscle LGE Imaging in the Detection of Acute Myocarditis
regions in these 3 patients (941  104 ms, more than
T1 mapping is similar to LGE and both significantly outperformed T2W imaging. AUC ¼ area
4 SD above the values of the normal cohort). The under the curve; CMR ¼ cardiac magnetic resonance; other abbreviations as in Figure 1.
finding of increased T1 values in skeletal muscle on
1054 Ferreira et al.
ShMOLLI T1 Mapping Detects Myocarditis
overall increase in T1 values in patient skeletal mus-
cle. All of these factors contribute to the under-
performance of dark-blood T2W imaging in the
detection of edema in a systemic disease such as acute
myocarditis.
Bright-blood T2W imaging. Newer bright-blood
T2W imaging offers some practical advantages:
breath-holds are shorter (typically under 10 s), and
the method is more forgiving in cases of tachyar-
rhythmias or patient movement. Better signal- and
contrast-to-noise ratios seemed to improve visuali-
zation of focal signal changes, especially in large
acute lesions such as ST-segment elevation my-
ocardial infarction (10); however, the detection of
small, focal patchy edema in acute myocarditis
was more challenging. When normal unaffected
myocardium could be reliably identified, bright-
blood T2W imaging tended to display these
changes well. Visual detection of global edema, on
the other hand, was less obvious to the eye on
bright-blood T2W images, and the use of skeletal
muscle as the reference region of interest can often
be limited by artifacts and noise over the region of
the muscle, in addition to the same issue sur-
rounding skeletal muscle inflammation as with
dark-blood T2W imaging discussed earlier in the
text (11). As previously demonstrated, for cases
of global edema, dark-blood T2W imaging had a
superior diagnostic performance over bright-blood
T2W imaging (11,19).
LGE imaging. LGE imaging has an excellent
contrast-to-noise ratio, making it one of the most
robust CMR methods for visualizing myocardial
lesions, including small focal disease. Acute
myocarditis can sometimes demonstrate little LGE
and/or predominantly global edema, which tends to
be inconspicuous on LGE imaging, such as in the
cases shown in Figure 3, hence the improved
diagnostic yield by combining any 2 of 3 CMR
methods for diagnosing the disease (7). It is also
conceivable that sometimes the lesions are so small,
they are beyond the resolution of even LGE
contrast imaging. As shown in Table 3, combining
LGE with T1 mapping did not significantly
improve the diagnostic performance of LGE alone,
but worsened the overall diagnostic performance of
T1 mapping alone; this may be related to the fact
that T1 mapping is able to detect changes caused by
both edema and myocyte necrosis, and thus can
serve to function as a “LGE þ T2W” combination,
only with much better sensitivity, diagnostic accu-
racy, and negative predictive value than any of the
LGE þ T2W combination set out in Table 3. From
a practical point of view, it may be possible for a T1
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013
OCTOBER 2013:1048–58
mapping sequence to be the only one required to
assess for all the criteria of acute myocarditis
(edema, hyperemia, and necrosis/scarring), serving
as the major component of a CMR myocarditis
protocol; this is worth exploring in future studies.
T1 mapping. T1 mapping using ShMOLLI cir-
cumvents most of the issues suffered by conven-
tional T2W and LGE imaging. It provides a direct,
quantitative means for myocardial characterization
without the need for contrast agents or reference
regions of interest to detect changes within the
myocardium. It has short breath-holds, is heart
rate independent, and is robust to even tachyar-
rhythmias, making it highly suitable for scanning
acutely ill patients. These advantages confer on
ShMOLLI T1 mapping its superior diagnostic
performance compared with the T2W and LGE
methods tested.
There may be a number of mechanisms that pro-
long myocardial T1 in acute myocarditis. Earlier
work had shown that myocardial T1 increased in
ischemic myocardium in canine models, which
largely, but not entirely, reflected the increase in water
content (20). It was postulated that T1 prolongation
may be due to changes in both total water content as
well as the relative amounts of water in intracellular
and extracellular space (20); in addition, it was also
hypothesized that changes in sodium and potassium
distribution may affect the motional freedom of
protons, contributing to prolongation of T1 values in
ischemic tissue. More recently, it has also been
demonstrated that T1 is significantly increased in
acute myocardial edema in animal (21) and human
(11) studies. Cellular edema, increased extracellular
space and water, inflammation, and myocyte necrosis
are common features of acute myocarditis (7,22), and
all of these pathophysiological processes may prolong
T1 values in the early stage.
T1 mapping had an excellent diagnostic perfor-
mance, with an w90% overall sensitivity, specificity,
and diagnostic accuracy for detecting changes in
myocarditis using a T1 cutoff of 990 ms. The selec-
tion of diagnostic thresholds depends on the amount
of noise in the technique and represents an exchange
between sensitivity and specificity. Historically, 2 SD
above the normal mean is accepted as a cutoff for
identifying abnormally high values. This was used,
for example, in the original validation of dark-blood
T2W imaging in detecting acute myocarditis (13)
and is commonly used in LGE imaging to identify
fibrosis in ischemic and nonischemic heart disease
(12,13,23). However, as shown in Table 3, the
threshold of 2 SD above the normal mean SI for
dark-blood T2W imaging is actually a relatively high
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013 Ferreira et al. 1055
OCTOBER 2013:1048–58 ShMOLLI T1 Mapping Detects Myocarditis

Figure 3. Three Cases of Acute Myocarditis With Global Edema but Only Mild LGE

CMR methods based on reference regions of interest (ROIs) significantly underestimate the areas of myocardial injury. Red indicates areas of
myocardium (myo) with values above the standard threshold for each CMR method. Green contour marks the positioning of myocardial reference
ROI. Myocardial contours are outlined in yellow. Skeletal muscle ROI not shown. TnI ¼ troponin I (mg/l); other abbreviations as in Figures 1 and 2.

threshold with good specificity but low sensitivity,
meaning that low-grade disease would easily be
missed. Interestingly, for T1 mapping, although the
threshold of T1 $ 990 ms is a sensitive threshold, it
simultaneously represents an even more stringent
>2.7 SD above normal T1 values. This is directly due
to the fact that normal myocardial T1 values have a
tight normal range with small variability in the
normal population (24), contributing to its good
diagnostic performance.

Figure 4. A Case of Acute Severe Myocarditis Demonstrating the Effect of Skeletal Muscle Inflammation on Dark-Blood T2W-CMR
Using Skeletal Muscle as a Reference Region to Detect Global Edema in the Myocardium

This patient had an initial ejection fraction of 37%, elevated C-reactive protein level of 94.6 mg/l (normal 0 to 8 mg/l), and a white blood cell
count 17.9  109/l, with recurrent ventricular tachycardia in hospital. (A) Dark-blood T2W imaging showed a visible increase in myocardial T2
SI. Inset: computer-aided analysis of the T2W image using skeletal muscle as a reference (red contour) failed to adequately detect a significant
increase in T2 SI ratio $2.0 relative to skeletal muscle within the myocardium (green contour). (B) T1 map of the corresponding slice in the
same patient. Purple contour (top right) outlines skeletal muscle (T1 ¼ 1,068  83 ms, arrow) corresponding to that on the T2W image; an
adjacent region of skeletal muscle (blue contour) also showed high T1 values (1,051  84 ms, arrow). Inset: computer-aided analysis of the T1
map, demonstrating global increase in the myocardial T1 value (1,221  157 ms). Abbreviations as in Figure 1.
1056 Ferreira et al.
ShMOLLI T1 Mapping Detects Myocarditis
Combination approach. In agreement with pub-
lished literature (7), a combination approach can
improve specificity to detect disease. On the basis of
the results of this study, T1 mapping has a superior
overall diagnostic performance over the T2W
techniques in the detection of acute myocarditis,
either alone or in combination with LGE; one may
argue that T1 mapping may be performed in place of
T2W imaging in select cases, although this
approach should be tested in larger studies, perhaps
comparing all currently available techniques in the
assessment of acute myocarditis. Although the
combination of T1 þ LGE was not significantly
better than LGE alone, on a practical level, the
main benefit of performing both T1 mapping and
LGE (rather than just LGE) would be to take
advantage of the superior sensitivity of T1 mapping
(90%) compared with LGE (and T2W imaging).
This may be especially useful when LGE is non-
revealing, when the predominant process is edema,
and in detecting subtle, small focal disease. It may
also be helpful if LGE imaging were not achieved
for reasons commonly encountered in the early
clinical setting (such as when the patient does not
tolerate the full protocol, has contraindications to
gadolinium, or lacks intravenous access).
Study limitations. This study was performed using
clinical validation for suspected myocarditis rather
than endomyocardial biopsy (EMB) as a reference
standard, and as such, lacks direct histopathological
confirmation of the diagnosis. However, in view of
the well-documented low sensitivity of EMB for
ruling out myocarditis (25–28) to serve as a reliable
gold standard and the fact that EMB is not routine
clinical practice in both hospitals, we had chosen to
validate the CMR methods using an internationally
accepted standard (7), as performed in multiple
previous CMR validation studies (7,13,18,29).
Although none of our patients underwent EMB, the
majority of the patients (92%) had a preceding sys-
temic viral or inflammatory illness, which, together
with their CMR findings, would be consistent with a
diagnosis of myocarditis; a few patients (n ¼ 4) did
not have a clear infectious or inflammatory precipi-
tant, but none of the 50 patients in the patient
cohort had an isolated ischemic pattern of LGE
suggestive of myocardial infarction as the presenting
diagnosis. Other obvious diagnoses, such as dilated
cardiomyopathy, hypertrophic cardiomyopathy, or
Takotsubo cardiomyopathy, were not included in
this study, as outlined in the Methods section.
Early gadolinium enhancement (EGE) was not
performed for practical reasons because our protocol
was already extensive. Therefore, of 3 available
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013
OCTOBER 2013:1048–58
conventional sequences, we selected dark-blood
T2W and LGE imaging because EGE is more
time consuming and its image quality often affected
by the irregular breathing patterns or arrhythmias
(7) commonly encountered in acutely ill patients.
Because the goal of the study was to validate novel
CMR techniques in a cohort defined by clinical
features, none of the Lake Louise imaging criteria
were used as inclusion criteria when enrolling sub-
jects. T2 mapping is another quantitative CMR
technique that has been shown to be useful in the
diagnosis of acute myocarditis (30) but was not
studied in this work. Future and larger studies may
directly compare all available CMR methods, such
as dark-blood T2, bright-blood T2, T2 mapping, T1
mapping, EGE, LGE, post-contrast T1, and
extracellular volume fraction mapping in order to
test method performance against each other and to
fully assess the utility of novel techniques in routine
clinical settings.
Because the accuracy of myocardial T1 measure-
ment remains to be established, no currently available
T1 mapping method can claim to accurately measure
T1. For T1 mapping methods based on inversion
recovery pulses, there may be systematic errors in T1
measurements related to inversion pulse efficiency
and other factors (31), and thus, T1 thresholds
established may be platform specific. Currently,
ShMOLLI appears to be an attractive method for T1
mapping because its known 4% underestimation of
T1 remains consistent over a wide range of native
tissue T1 (13); it also has a stable, narrow range of
normal myocardial T1 values (24), making it highly
suitable for detecting disease states (11,12,32–34).
Increase in native myocardial T1 values is
nonspecific and is seen in a number of myocardial
diseases other than acute myocardial injury, including
cardiac amyloidosis (32), hypertrophic and dilated
cardiomyopathy (34), diffuse fibrosis (33), and
chronic infarct scars (12,35), and thus, as with all
other diagnostic tools, must be interpreted within the
clinical context. The exact mechanism(s) leading to
prolonged T1 values in different cardiac diseases
remain to be established.
CONCLUSIONS
T1 mapping and bright-blood T2W-CMR are
novel methods sensitive to the detection of acute
myocarditis. T1 mapping showed excellent and su-
perior diagnostic performance compared with
T2W-CMR, with a high sensitivity compared with
T2W and LGE techniques, which may be especially
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013 Ferreira et al. 1057
OCTOBER 2013:1048–58 ShMOLLI T1 Mapping Detects Myocarditis

useful in detecting subtle focal disease and in cases and Professors Adrian Banning and Keith
where gadolinium contrast imaging is not feasible. Channon.

Acknowledgments
The authors acknowledge contributions from the
interventional cardiology team at the John Radcliffe
Hospital, including Drs. Nicholas Alp, Colin For-
far, Rajesh Kharbanda, and Bernard Prendergast,
Reprint requests and correspondence: Dr. Stefan
Neubauer, Department of Cardiovascular Medicine,
University of Oxford, John Radcliffe Hospital, Oxford
OX3 9DU, United Kingdom. E-mail: stefan.neubauer@
cardiov.ox.ac.uk.

REFERENCES

1. Monney PA, Sekhri N, Burchell T,
et al. Acute myocarditis presenting as
acute coronary syndrome: role of early
cardiac magnetic resonance in its
diagnosis. Heart 2010;97:1312–8.
2. Assomull RG, Lyne JC, Keenan N,
et al. The role of cardiovascular mag-
netic resonance in patients presenting
with chest pain, raised troponin, and
unobstructed coronary arteries. Eur
Heart J 2007;28:1242–9.
3. Baccouche H, Mahrholdt H,
Meinhardt G, et al. Diagnostic synergy
of non-invasive cardiovascular mag-
netic resonance and invasive endo-
myocardial biopsy in troponin-positive
patients without coronary artery dis-
ease. Eur Heart J 2009;30:2869–79.
4. Gallagher S, Jones DA, Anand V,
Mohiddin S. Diagnosis and manage-
ment of patients with acute cardiac
symptoms, troponin elevation and
culprit-free angiograms. Heart 2012;
98:974–81.
5. Doolan A, Langlois N, Semsarian C.
Causes of sudden cardiac death in
young Australians. Med J Aust 2004;
180:110–2.
6. Felker GM, Thompson RE, Hare JM,
et al. Underlying causes and long-term
survival in patients with initially un-
explained cardiomyopathy. N Engl J
Med 2000;342:1077–84.
7. Friedrich MG, Sechtem U, Schulz-
Menger J, et al. Cardiovascular mag-
netic resonance in myocarditis: a
JACC white paper. J Am Coll Cardiol
2009;53:1475–87.
8. Kellman P, Aletras AH, Mancini C,
McVeigh ER, Arai AE. T2-prepared
SSFP improves diagnostic confidence
in edema imaging in acute myocardial
infarction compared to turbo spin echo.
Magn Reson Med 2007;57:891–7.
9. Aletras AH, Kellman P, Derbyshire JA,
Arai AE. ACUT2E TSE-SSFP: a
hybrid method for T2-weighted imag-
ing of edema in the heart. Magn Reson
Med 2008;59:229–35.
10. Payne AR, Casey M, McClure J, et al.
Bright-blood T2-weighted MRI has
higher diagnostic accuracy than dark-
blood short tau inversion recovery
MRI for detection of acute myocardial
infarction and for assessment of the
ischemic area at risk and myocardial
salvage. Circ Cardiovasc Imaging
2011;4:210–9.
11. Ferreira V, Piechnik S, Dall’Armellina E,
et al. Non-contrast T1-mapping detects
acute myocardial edema with high
diagnostic accuracy: a comparison to
T2-weighted cardiovascular magnetic
resonance. J Cardiovasc Magn Reson
2012;14:42.
12. Dall’Armellina E, Piechnik S,
Ferreira V, et al. Cardiovascular mag-
netic resonance by non contrast T1
mapping allows assessment of severity
of injury in acute myocardial infarction.
J Cardiovasc Magn Reson 2012;14:15.
13. Abdel-Aty H, Boye P, Zagrosek A,
et al. Diagnostic performance of car-
diovascular magnetic resonance in pa-
tients with suspected acute myocarditis:
comparison of different approaches.
J Am Coll Cardiol 2005;45:1815–22.
14. Piechnik SK, Ferreira VM,
Dall’Armellina E, et al. Shortened
Modified Look-Locker Inversion re-
covery (ShMOLLI) for clinical
myocardial T1-mapping at 1.5 and 3
T within a 9 heartbeat breathhold.
J Cardiovasc Magn Reson 2010;12:69.
15. Simonetti OP, Finn JP, White RD,
Laub G, Henry DA. “Black blood”
T2-weighted inversion-recovery MR
imaging of the heart. Radiology 1996;
199:49–57.
16. Kellman P, Arai AE, McVeigh ER,
Aletras AH. Phase-sensitive inversion
recovery for detecting myocardial
infarction using gadolinium-delayed
hyperenhancement. Magn Reson
Med 2002;47:372–83.
17. DeLong ER, DeLong DM, Clarke-
Pearson DL. Comparing the areas
under two or more correlated receiver
operating characteristic curves: a
nonparametric approach. Biometrics
1988;44:837–45.
18. Laissy JP, Messin B, Varenne O, et al.
MRI of acute myocarditis: a compre-
hensive approach based on various
imaging sequences. Chest 2002;122:
1638–48.
19. O h-Ici D, Ridgway J, Kuehne T, et al.
Cardiovascular magnetic resonance of
myocardial edema using a short
inversion time inversion recovery
(STIR) black-blood technique: Diag-
nostic accuracy of visual and semi-
quantitative assessment. J Cardiovasc
Magn Reson 2012;14:22.
20. Williams ES, Kaplan JI, Thatcher F.
Prolongation of proton spin lattice
relaxation times in regionally ischemic
tissue from dog hearts. J Nucl Med
1980;21:449–53.
21. Ugander M, Bagi PS, Oki AJ, et al.
Myocardial edema as detected by
pre-contrast T1 and T2 CMR de-
lineates area at risk associated with
acute myocardial infarction. J Am Coll
Cardiol Img 2012;5:596–603.
22. Liu PP, Mason JW. Advances in the
understanding of myocarditis. Circu-
lation 2001;104:1076–82.
23. Kim RJ, Fieno DS, Parrish TB, et al.
Relationship of MRI delayed contrast
enhancement to irreversible injury,
infarct age, and contractile function.
Circulation 1999;100:1992–2002.
24. Piechnik S, Ferreira V, Lewandowski A,
et al. Normal variation of magnetic
resonance T1 relaxation times in the
human population at 1.5T using
ShMOLLI. J Cardiovasc Magn Reson
2013;15:13.
25. Baughman KL. Diagnosis of myocar-
ditis death of Dallas criteria. Circula-
tion 2006;113:593–5.
26. Hauck AJ, Kearney DL, Edwards WD.
Evaluation of postmortem endomyo-
cardial biopsy specimens from 38 pa-
tients with lymphocytic myocarditis:
implications for role of sampling error.
Mayo Clin Proc 1989;64:1235–45.
27. Mavrogeni S, Spargias C, Bratis C,
et al. Myocarditis as a precipitating
factor for heart failure: evaluation and
1-year follow-up using cardiovascular
magnetic resonance and endomyo-
cardial biopsy. Eur J Heart Fail 2011;
13:830–7.
28. Mahrholdt H, Goedecke C, Wagner A,
et al. Cardiovascular magnetic reso-
nance assessment of human myocar-
ditis: a comparison to histology and
1058 Ferreira et al.
ShMOLLI T1 Mapping Detects Myocarditis
molecular pathology. Circulation 2004;
109:1250–8.
29. Friedrich MG, Strohm O, Schulz-
Menger J, Marciniak H, Luft FC,
Dietz R. Contrast media-enhanced
magnetic resonance imaging visualizes
myocardial changes in the course of
viral myocarditis. Circulation 1998;97:
1802–9.
30. Thavendiranathan P, Walls M, Giri S,
et al. Improved detection of myocar-
dial involvement in acute inflamma-
tory cardiomyopathies using T2
mapping. Circ Cardiovasc Imaging
2011;5:102–10.
31. Rodgers CT, Piechnik SK, Delabarre LJ,
et al. Inversion recovery at 7 T in the
human myocardium: measurement of
T(1), inversion efficiency and B(1) (þ).
JACC: CARDIOVASCULAR IMAGING, VOL. 6, NO. 10, 2013
OCTOBER 2013:1048–58
Magn Reson Med 2012 Nov 29 [E-pub 35. Dall’Armellina E, Ferreira V,
ahead of print]. Kharbanda R, et al. Diagnostic value of
pre-contrast T1 mapping in acute and
32. Karamitsos TD, Piechnik SK,
chronic myocardial infarction. J Am
Banypersad S, et al. Non-contrast T1 Coll Cardiol Img 2013;6:739–42.
mapping for the diagnosis of cardiac
amyloidosis. J Am Coll Cardiol Img
2013;6:488–97.
33. Bull S, White SK, Piechnik SK, et al. Key Words: cardiac magnetic
Human non-contrast T1 values and resonance - myocarditis -
correlation with histology in diffuse
fibrosis. Heart 2013;99:932–7. ShMOLLI - T1 mapping -
34. Dass S, Suttie JJ, Piechnik SK, et al.
T2-weighted CMR.
Myocardial tissue characterization us-
ing magnetic resonance noncontrast
T1 mapping in hypertrophic and dilated
APPENDIX
cardiomyopathy / clinical perspective.
Circ Cardiovasc Imaging 2012;5: For an expanded Methods section, please see
726–33. the online version of this paper.